TheMatAllen

Karma: 134

TheMatAllen 26 Oct 2023 4:23 UTC
1 point
0 ∶ 0
on: EA Data Science—Community call
The slack link to EA-Anywhere is expired

TheMatAllen 10 Oct 2023 19:28 UTC
3 points
1 ∶ 0
in reply to: AlanGreenspan’s comment on: Debate series: should we push for a pause on the development of AI?
I’m surprised no one commented in response.

My initial impression is that if 75% of the world was working on stopping AI, that would presumably include government and policy who could very well tell the remaining 25% that they can’t work on AI.

As far the infohazards of thinking about developing AI; I think that’s beyond what we can do. Ada Lovelace thought about AI in the 1840′s (https://www.nist.gov/blogs/taking-measure/ada-lovelace-worlds-first-computer-programmer-who-predicted-artificial), and the possibility of massive economic gains is enough to encourage even a small number of people to continue working on it (and if fewer people are thinking about it, the competition is even smaller, thus the rewards much greater).

I do agree that there are awful things going on in the world, and ideally AI will be able to solve or massively fix those problems (on a scale that humans just aren’t able to).

TheMatAllen 1 Oct 2023 5:28 UTC
2 points
0 ∶ 0
in reply to: Bryce Robertson’s comment on: Anki deck for learning the main AI safety orgs, projects, and programs
Interesting, I’d be happy to give you feedback on how I like the cards (I have a reminder for ~2 months from now). For transparency, my initial goal is to have a better sense of all orgs out in the ecosystem and have it in a form that I can slow-feed to myself, not necessarily memorize these for the rest of my life.

Maybe the people you talked to didn’t think having these memorized would provide benefit? Not sure if they were already fans of SRS before. I’d argue that having even a “Fermi estimate” equivalent of what people are doing and what options are out there have benefits for knowing what to apply to, what to suggest to others, etc.

I can’t promise that my cards are that high quality! I almost only use cloze deletions and that may not be preferable. I do source all of my links and the source text that I’m roughly basing my card on. If you’re open to taking a look at them it’d be great to get quick feedback on whether they’re helpful or not, and especially whether it’d be useful to put more work into for a deck of the entire project.

TheMatAllen 30 Sep 2023 1:28 UTC
3 points
0 ∶ 0
on: Anki deck for learning the main AI safety orgs, projects, and programs
Thank you for making this deck! I’m an avid Anki user and will start on this today.

I’m curious about how useful decks like these would be for people. I’m going through AGISF right now and making a bunch of cards for studying and the thought of publishing them for others came up, but wasn’t sure if it was worth the effort (to polish them for public use).

@Bryce Robertson, any thoughts on this? Were you approached to do this or did you come up with your own reasons as to why you started this project? If this is something that would be valuable for other resources I’d be quite excited to work on this.

TheMatAllen 5 Feb 2023 15:24 UTC
1 point
0 ∶ 0
in reply to: Arjun Panickssery’s comment on: EAGxBerkeley 2022 Retrospective
From the survey and mentions directly to me

TheMatAllen 26 Dec 2022 20:00 UTC
7 points
1 ∶ 0
on: Savoring my moral circle
Reminds me of the neologism, “sonder”

TheMatAllen 11 Oct 2022 20:21 UTC
4 points
1 ∶ 0
on: On absurdity
This reminded me of this quote from Victor Frankl, “The attempt to develop a sense of humor and to see things in a humorous light is some kind of a trick learned while mastering the art of living.”

When looking it up here, the tags made me laugh; “tags: holocaust, humor, suffering”

TheMatAllen 20 May 2022 21:16 UTC
1 point
0 ∶ 0
in reply to: Rob Mitchell’s comment on: Thoughts on requesting reasoning or examples to not pursue fields/positions
“Decide in full knowledge”; I think that’s exactly what I was aiming for, thank you.

My impression (and maybe I just haven’t seen the articles) is that there is more focus on “these are the traits of people who have done well in this path” and something along these lines would attempt to balance that.

I may also be biased as I have a lot of interests, and narrowing down feels much better than adding more options to my plate.

TheMatAllen 4 Mar 2022 1:20 UTC
1 point
0 ∶ 0
on: Learning By Writing
Thanks for this post! I especially enjoy your emphasis on creating a hypothesis as a “flag pole” that you can iteratively test your understanding against; I imagine this is a great way to embody Kant’s advice of “Have the courage to use your own reason” as well as getting used to overcoming Confirmation Bias. Really cool suggestion that I’ll be implementing, thank you!
It seems there is some interest in the public scrutiny of ideas when going through this type of research (from Jeremy, David B, and Laain) and I’d also enjoy having a space of idea scrutiny to get that feedback from outside. Would you all be open to an initial low investment testing of a space like this?
If we have common areas we’d like to investigate we can give each other our feedback (Step 4 and 8) to reduce the individual strain that Holden describes above. Maybe a lite test of this would be to have a small email thread bringing attention to topics that feedback is being asked on, and another area (slack/discord?) that allows the hashing out of said claims? I believe this could increase the group’s familiarity with the work as well as reduce the strain of an individual. If this could be done at scale, I initially believe this could serve as a great method of community updating and involvement. Finally, this exploration is certainly acting in the “not a finished thought, but am discovering an answer” and am open to ideas.

TheMatAllen 23 Feb 2022 4:49 UTC
1 point
0 ∶ 0
in reply to: Jeremy’s comment on: Low-Commitment Less Wrong Book (EG Article) Club
Ah, I see. Well, I’m all in for the lower commitment at this point. Will you be advertising the start on this post?

TheMatAllen 22 Feb 2022 4:53 UTC
2 points
0 ∶ 0
on: Low-Commitment Less Wrong Book (EG Article) Club
I’m interested as well, for participating or helping with organization.
Idea when project is up to speed; if multiple groups are operating at the same time, have a central theme that all each group focuses on and at the end of a “sequence” of these a larger article could be written on the collective insights developed. Something along those lines.